)]}'
{"/COMMIT_MSG":[{"author":{"_account_id":7249,"name":"Ann Taraday","email":"akamyshnikova@mirantis.com","username":"AKamyshnikova"},"change_message_id":"c957608de914d89532dfca6645c945ff21d25c43","unresolved":false,"context_lines":[{"line_number":19,"context_line":"identical to that of Nova, we can lower the number of"},{"line_number":20,"context_line":"greenlets per process to be more in line with a typical"},{"line_number":21,"context_line":"max database connection pool size."},{"line_number":22,"context_line":""},{"line_number":23,"context_line":"Change-Id: I94cd2f9262e0f330cf006b40bb3c0071086e5d71"}],"source_content_type":"text/x-gerrit-commit-message","patch_set":4,"id":"3a57f1b5_c609f7de","line":22,"updated":"2016-02-11 11:41:57.000000000","message":"I think you should add DocImpact.","commit_id":"0cb5c751833752f681a80e60ecf3ebc6addcdd79"},{"author":{"_account_id":11816,"name":"mike bayer","display_name":"Mike Bayer","email":"mike_mp@zzzcomputing.com","username":"zzzeek","status":"Red Hat"},"change_message_id":"668fef9ea903c66c30d25601432564919223664b","unresolved":false,"context_lines":[{"line_number":19,"context_line":"identical to that of Nova, we can lower the number of"},{"line_number":20,"context_line":"greenlets per process to be more in line with a typical"},{"line_number":21,"context_line":"max database connection pool size."},{"line_number":22,"context_line":""},{"line_number":23,"context_line":"Change-Id: I94cd2f9262e0f330cf006b40bb3c0071086e5d71"}],"source_content_type":"text/x-gerrit-commit-message","patch_set":4,"id":"dae33548_95ab4b89","line":22,"in_reply_to":"3a57f1b5_c609f7de","updated":"2016-02-15 15:39:07.000000000","message":"can you show me an example ?","commit_id":"0cb5c751833752f681a80e60ecf3ebc6addcdd79"},{"author":{"_account_id":11816,"name":"mike bayer","display_name":"Mike Bayer","email":"mike_mp@zzzcomputing.com","username":"zzzeek","status":"Red Hat"},"change_message_id":"638f52eaf3e7f6e390f15d49d019106b278227ca","unresolved":false,"context_lines":[{"line_number":19,"context_line":"identical to that of Nova, we can lower the number of"},{"line_number":20,"context_line":"greenlets per process to be more in line with a typical"},{"line_number":21,"context_line":"max database connection pool size."},{"line_number":22,"context_line":""},{"line_number":23,"context_line":"Change-Id: I94cd2f9262e0f330cf006b40bb3c0071086e5d71"}],"source_content_type":"text/x-gerrit-commit-message","patch_set":4,"id":"dae33548_f5f9f74e","line":22,"in_reply_to":"dae33548_95ab4b89","updated":"2016-02-15 15:46:02.000000000","message":"Done","commit_id":"0cb5c751833752f681a80e60ecf3ebc6addcdd79"},{"author":{"_account_id":748,"name":"Armando Migliaccio","email":"armamig@gmail.com","username":"armando-migliaccio"},"change_message_id":"ebe12218eb4d8ebe28e4014630875d91df1d0fab","unresolved":false,"context_lines":[{"line_number":20,"context_line":"greenlets per process to be more in line with a typical"},{"line_number":21,"context_line":"max database connection pool size."},{"line_number":22,"context_line":""},{"line_number":23,"context_line":"DocImpact"},{"line_number":24,"context_line":"Change-Id: I94cd2f9262e0f330cf006b40bb3c0071086e5d71"}],"source_content_type":"text/x-gerrit-commit-message","patch_set":6,"id":"dae33548_b1793eff","line":23,"updated":"2016-02-15 17:25:42.000000000","message":"for documentation rules, DocImpact should be followed by a brief description of the change.","commit_id":"f7ef174e27b376574d9f2445c8f3c603e794c770"},{"author":{"_account_id":11816,"name":"mike bayer","display_name":"Mike Bayer","email":"mike_mp@zzzcomputing.com","username":"zzzeek","status":"Red Hat"},"change_message_id":"ea55e3cb0fecb7d8d4a9f43a54573e340a0376a4","unresolved":false,"context_lines":[{"line_number":20,"context_line":"greenlets per process to be more in line with a typical"},{"line_number":21,"context_line":"max database connection pool size."},{"line_number":22,"context_line":""},{"line_number":23,"context_line":"DocImpact"},{"line_number":24,"context_line":"Change-Id: I94cd2f9262e0f330cf006b40bb3c0071086e5d71"}],"source_content_type":"text/x-gerrit-commit-message","patch_set":6,"id":"3afc51ec_eb4928d2","line":23,"in_reply_to":"dae33548_3fc21534","updated":"2016-03-11 19:04:07.000000000","message":"done","commit_id":"f7ef174e27b376574d9f2445c8f3c603e794c770"},{"author":{"_account_id":11816,"name":"mike bayer","display_name":"Mike Bayer","email":"mike_mp@zzzcomputing.com","username":"zzzeek","status":"Red Hat"},"change_message_id":"7677d2c23f9268d11f56835f52c239fc019c98b1","unresolved":false,"context_lines":[{"line_number":20,"context_line":"greenlets per process to be more in line with a typical"},{"line_number":21,"context_line":"max database connection pool size."},{"line_number":22,"context_line":""},{"line_number":23,"context_line":"DocImpact"},{"line_number":24,"context_line":"Change-Id: I94cd2f9262e0f330cf006b40bb3c0071086e5d71"}],"source_content_type":"text/x-gerrit-commit-message","patch_set":6,"id":"dae33548_3fc21534","line":23,"in_reply_to":"dae33548_b1793eff","updated":"2016-02-15 18:01:47.000000000","message":"is that formatted as:\n\nDocImpact: \u003cbrief description\u003e \n\n?","commit_id":"f7ef174e27b376574d9f2445c8f3c603e794c770"},{"author":{"_account_id":9656,"name":"Ihar Hrachyshka","email":"ihrachys@redhat.com","username":"ihrachys","status":"Red Hat Networking Systems Engineer"},"change_message_id":"14346a11e911958d9166d1ad9fe2f1a7b1cd6357","unresolved":false,"context_lines":[{"line_number":20,"context_line":"greenlets per process to be more in line with a typical"},{"line_number":21,"context_line":"max database connection pool size."},{"line_number":22,"context_line":""},{"line_number":23,"context_line":"DocImpact: a new configuration value wsgi_default_pool_size is added"},{"line_number":24,"context_line":"Change-Id: I94cd2f9262e0f330cf006b40bb3c0071086e5d71"}],"source_content_type":"text/x-gerrit-commit-message","patch_set":9,"id":"da0c15f0_ca4b8158","line":23,"updated":"2016-03-24 23:15:12.000000000","message":"technically, it\u0027s not correct: the option was present for a while, it just was not used for neutron-server.","commit_id":"a984081d9d33476be92af432fa8f19b863607de5"},{"author":{"_account_id":11816,"name":"mike bayer","display_name":"Mike Bayer","email":"mike_mp@zzzcomputing.com","username":"zzzeek","status":"Red Hat"},"change_message_id":"6924f599049f747cde0739d817c6fb1b436321fb","unresolved":false,"context_lines":[{"line_number":20,"context_line":"greenlets per process to be more in line with a typical"},{"line_number":21,"context_line":"max database connection pool size."},{"line_number":22,"context_line":""},{"line_number":23,"context_line":"DocImpact: a new configuration value wsgi_default_pool_size is added"},{"line_number":24,"context_line":"Change-Id: I94cd2f9262e0f330cf006b40bb3c0071086e5d71"}],"source_content_type":"text/x-gerrit-commit-message","patch_set":9,"id":"9a061dce_531cee76","line":23,"in_reply_to":"da0c15f0_ca4b8158","updated":"2016-04-06 02:34:58.000000000","message":"Done","commit_id":"a984081d9d33476be92af432fa8f19b863607de5"}],"neutron/wsgi.py":[{"author":{"_account_id":6681,"name":"Brent Eagles","email":"beagles@redhat.com","username":"beagles"},"change_message_id":"dafa2ccf03c2a1118a1d5675b2532941787e611f","unresolved":false,"context_lines":[{"line_number":57,"context_line":"                help\u003d_(\u0027Enable SSL on the API server\u0027)),"},{"line_number":58,"context_line":"    cfg.IntOpt(\u0027wsgi_default_pool_size\u0027,"},{"line_number":59,"context_line":"               default\u003d1000,"},{"line_number":60,"context_line":"               help\u003d\"Size of the pool of greenthreads used by wsgi\"),"},{"line_number":61,"context_line":""},{"line_number":62,"context_line":"]"},{"line_number":63,"context_line":""}],"source_content_type":"text/x-python","patch_set":1,"id":"3a57f1b5_937ba67b","line":60,"updated":"2016-02-09 18:50:00.000000000","message":"While ultimately this controls the size of the thread pool used by eventlet.wsgi  when starting the server, would it be more informative to state that this actually constrains the number of client connections for this server?","commit_id":"c71bb0155109610783c255389c641d9a3c1ce72e"},{"author":{"_account_id":11816,"name":"mike bayer","display_name":"Mike Bayer","email":"mike_mp@zzzcomputing.com","username":"zzzeek","status":"Red Hat"},"change_message_id":"384a3cb46910f4ebb3ccf14cb34cbfbb9906c844","unresolved":false,"context_lines":[{"line_number":57,"context_line":"                help\u003d_(\u0027Enable SSL on the API server\u0027)),"},{"line_number":58,"context_line":"    cfg.IntOpt(\u0027wsgi_default_pool_size\u0027,"},{"line_number":59,"context_line":"               default\u003d1000,"},{"line_number":60,"context_line":"               help\u003d\"Size of the pool of greenthreads used by wsgi\"),"},{"line_number":61,"context_line":""},{"line_number":62,"context_line":"]"},{"line_number":63,"context_line":""}],"source_content_type":"text/x-python","patch_set":1,"id":"3a57f1b5_b5521112","line":60,"in_reply_to":"3a57f1b5_937ba67b","updated":"2016-02-10 17:35:00.000000000","message":"I don\u0027t know that this is true for Neutron.  It isn\u0027t true for Nova, because the greenlet pool is local to a single process; Nova spawns as many processes as there are cores on the machine.   We\u0027ve observed that Neutron spins up lots of processes as well, so assuming those are all handling individual greenlet pools, then the number of client connections is much larger than this number.  Also, when Neutron etc. configure the socket connection, there\u0027s a \"backlog\" setting which here seems to be 4096, which means a client isn\u0027t rejected, but has to wait, that would impact this number as well.","commit_id":"c71bb0155109610783c255389c641d9a3c1ce72e"},{"author":{"_account_id":6681,"name":"Brent Eagles","email":"beagles@redhat.com","username":"beagles"},"change_message_id":"65a1862ccf2ff1ac70d452fda4228e348f585767","unresolved":false,"context_lines":[{"line_number":57,"context_line":"                help\u003d_(\u0027Enable SSL on the API server\u0027)),"},{"line_number":58,"context_line":"    cfg.IntOpt(\u0027wsgi_default_pool_size\u0027,"},{"line_number":59,"context_line":"               default\u003d1000,"},{"line_number":60,"context_line":"               help\u003d\"Size of the pool of greenthreads used by wsgi\"),"},{"line_number":61,"context_line":""},{"line_number":62,"context_line":"]"},{"line_number":63,"context_line":""}],"source_content_type":"text/x-python","patch_set":1,"id":"3a57f1b5_7ba9e37a","line":60,"in_reply_to":"3a57f1b5_b5521112","updated":"2016-02-10 19:19:29.000000000","message":"Ah, right. I guess it could be \"the number of active connections per API worker\", but it\u0027s just a thought.","commit_id":"c71bb0155109610783c255389c641d9a3c1ce72e"},{"author":{"_account_id":6681,"name":"Brent Eagles","email":"beagles@redhat.com","username":"beagles"},"change_message_id":"dafa2ccf03c2a1118a1d5675b2532941787e611f","unresolved":false,"context_lines":[{"line_number":121,"context_line":"    def __init__(self, name, num_threads\u003dNone):"},{"line_number":122,"context_line":"        # Raise the default from 8192 to accommodate large tokens"},{"line_number":123,"context_line":"        eventlet.wsgi.MAX_HEADER_LINE \u003d CONF.max_header_line"},{"line_number":124,"context_line":"        self.num_threads \u003d num_threads or self.default_pool_size"},{"line_number":125,"context_line":"        # Pool for a greenthread in which wsgi server will be running"},{"line_number":126,"context_line":"        self.pool \u003d eventlet.GreenPool(1)"},{"line_number":127,"context_line":"        self.name \u003d name"}],"source_content_type":"text/x-python","patch_set":1,"id":"3a57f1b5_b8a3214f","line":124,"updated":"2016-02-09 18:50:00.000000000","message":"Out of curiousity, why the intermediate class variable? Why not just CONF.wsgi_default_pool_size here?","commit_id":"c71bb0155109610783c255389c641d9a3c1ce72e"},{"author":{"_account_id":11816,"name":"mike bayer","display_name":"Mike Bayer","email":"mike_mp@zzzcomputing.com","username":"zzzeek","status":"Red Hat"},"change_message_id":"384a3cb46910f4ebb3ccf14cb34cbfbb9906c844","unresolved":false,"context_lines":[{"line_number":121,"context_line":"    def __init__(self, name, num_threads\u003dNone):"},{"line_number":122,"context_line":"        # Raise the default from 8192 to accommodate large tokens"},{"line_number":123,"context_line":"        eventlet.wsgi.MAX_HEADER_LINE \u003d CONF.max_header_line"},{"line_number":124,"context_line":"        self.num_threads \u003d num_threads or self.default_pool_size"},{"line_number":125,"context_line":"        # Pool for a greenthread in which wsgi server will be running"},{"line_number":126,"context_line":"        self.pool \u003d eventlet.GreenPool(1)"},{"line_number":127,"context_line":"        self.name \u003d name"}],"source_content_type":"text/x-python","patch_set":1,"id":"3a57f1b5_b5ee9120","line":124,"in_reply_to":"3a57f1b5_b8a3214f","updated":"2016-02-10 17:35:00.000000000","message":"I copied what Nova does, as I don\u0027t have deep knowledge of oslo.config I\u0027m not sure if there\u0027s a real reason for this.  Perhaps they are trying to guard against test suites etc. that modify CONF, not really sure.","commit_id":"c71bb0155109610783c255389c641d9a3c1ce72e"},{"author":{"_account_id":6524,"name":"Henry Gessau","email":"HenryG@gessau.net","username":"gessau"},"change_message_id":"4ea635f5f898d8bae067c4cf35dcb8d760ccb20c","unresolved":false,"context_lines":[{"line_number":116,"context_line":"class Server(object):"},{"line_number":117,"context_line":"    \"\"\"Server class to manage multiple WSGI sockets and applications.\"\"\""},{"line_number":118,"context_line":""},{"line_number":119,"context_line":"    default_pool_size \u003d CONF.wsgi_default_pool_size"},{"line_number":120,"context_line":""},{"line_number":121,"context_line":"    def __init__(self, name, num_threads\u003dNone):"},{"line_number":122,"context_line":"        # Raise the default from 8192 to accommodate large tokens"}],"source_content_type":"text/x-python","patch_set":2,"id":"3a57f1b5_dbb6d789","line":119,"updated":"2016-02-10 19:27:11.000000000","message":"I understand you copied what nova does, but I say we dispense of this class variable for Neutron until we have a good reason for it.","commit_id":"1bc5bcc81e5a31120b11666392e9a5abd6169bc9"},{"author":{"_account_id":11816,"name":"mike bayer","display_name":"Mike Bayer","email":"mike_mp@zzzcomputing.com","username":"zzzeek","status":"Red Hat"},"change_message_id":"6b6c649b9d09bfba29d01ea1ff4ea8bc64e86fcc","unresolved":false,"context_lines":[{"line_number":116,"context_line":"class Server(object):"},{"line_number":117,"context_line":"    \"\"\"Server class to manage multiple WSGI sockets and applications.\"\"\""},{"line_number":118,"context_line":""},{"line_number":119,"context_line":"    default_pool_size \u003d CONF.wsgi_default_pool_size"},{"line_number":120,"context_line":""},{"line_number":121,"context_line":"    def __init__(self, name, num_threads\u003dNone):"},{"line_number":122,"context_line":"        # Raise the default from 8192 to accommodate large tokens"}],"source_content_type":"text/x-python","patch_set":2,"id":"3a57f1b5_c2a4a3a7","line":119,"in_reply_to":"3a57f1b5_dbb6d789","updated":"2016-02-10 21:52:08.000000000","message":"done, I only care about the names being consistent on the outside.","commit_id":"1bc5bcc81e5a31120b11666392e9a5abd6169bc9"},{"author":{"_account_id":10061,"name":"Joe Talerico","email":"jtaleric@redhat.com","username":"rook"},"change_message_id":"023df590e573421b72b127f9db90f1c42a1c6cd1","unresolved":false,"context_lines":[{"line_number":57,"context_line":"                help\u003d_(\u0027Enable SSL on the API server\u0027)),"},{"line_number":58,"context_line":"    cfg.IntOpt(\u0027wsgi_default_pool_size\u0027,"},{"line_number":59,"context_line":"               default\u003d1000,"},{"line_number":60,"context_line":"               help\u003d\"Size of the pool of greenthreads used by wsgi\"),"},{"line_number":61,"context_line":""},{"line_number":62,"context_line":"]"},{"line_number":63,"context_line":""}],"source_content_type":"text/x-python","patch_set":3,"id":"3a57f1b5_023cfbe0","line":60,"updated":"2016-02-10 21:55:28.000000000","message":"help should be _(\"\")","commit_id":"75b1df8616eef3f68f0ad61f940ace91e1cec770"},{"author":{"_account_id":9515,"name":"Matthew Kassawara","email":"mkassawara@gmail.com","username":"ionosphere80"},"change_message_id":"91d9c4fcd5b3d94a82cd98c2c188b82c8ebdf174","unresolved":false,"context_lines":[{"line_number":57,"context_line":"                help\u003d_(\u0027Enable SSL on the API server\u0027)),"},{"line_number":58,"context_line":"    cfg.IntOpt(\u0027wsgi_default_pool_size\u0027,"},{"line_number":59,"context_line":"               default\u003d1000,"},{"line_number":60,"context_line":"               help\u003d_(\"Size of the pool of greenthreads used by wsgi\")),"},{"line_number":61,"context_line":""},{"line_number":62,"context_line":"]"},{"line_number":63,"context_line":""}],"source_content_type":"text/x-python","patch_set":5,"id":"dae33548_d17322b0","line":60,"updated":"2016-02-15 17:44:11.000000000","message":"Suggestion: Size of WSGI thread pool. Default value of 1000 accounts for typical maximum size of database connection pool.","commit_id":"f2db135c9349706b4c85b9eb6b5bab51373773f9"},{"author":{"_account_id":11816,"name":"mike bayer","display_name":"Mike Bayer","email":"mike_mp@zzzcomputing.com","username":"zzzeek","status":"Red Hat"},"change_message_id":"b2383190e6bb2e895e65fa6bc873517463198d54","unresolved":false,"context_lines":[{"line_number":57,"context_line":"                help\u003d_(\u0027Enable SSL on the API server\u0027)),"},{"line_number":58,"context_line":"    cfg.IntOpt(\u0027wsgi_default_pool_size\u0027,"},{"line_number":59,"context_line":"               default\u003d1000,"},{"line_number":60,"context_line":"               help\u003d_(\"Size of the pool of greenthreads used by wsgi\")),"},{"line_number":61,"context_line":""},{"line_number":62,"context_line":"]"},{"line_number":63,"context_line":""}],"source_content_type":"text/x-python","patch_set":5,"id":"dae33548_df35e946","line":60,"in_reply_to":"dae33548_d17322b0","updated":"2016-02-15 18:03:44.000000000","message":"well the default of the connection pool is 15, neutron seems to bump it to 30 in neutron/common/config.py.   1000 is just way too big so I think a help message like that would be misleading.","commit_id":"f2db135c9349706b4c85b9eb6b5bab51373773f9"},{"author":{"_account_id":11816,"name":"mike bayer","display_name":"Mike Bayer","email":"mike_mp@zzzcomputing.com","username":"zzzeek","status":"Red Hat"},"change_message_id":"049bff3299b05a70e26b86a583a10a12918d15cf","unresolved":false,"context_lines":[{"line_number":57,"context_line":"                help\u003d_(\u0027Enable SSL on the API server\u0027)),"},{"line_number":58,"context_line":"    cfg.IntOpt(\u0027wsgi_default_pool_size\u0027,"},{"line_number":59,"context_line":"               default\u003d1000,"},{"line_number":60,"context_line":"               help\u003d_(\"Size of the pool of greenthreads used by wsgi\")),"},{"line_number":61,"context_line":""},{"line_number":62,"context_line":"]"},{"line_number":63,"context_line":""}],"source_content_type":"text/x-python","patch_set":5,"id":"3afc51ec_2b2b10ed","line":60,"in_reply_to":"dae33548_df35e946","updated":"2016-03-11 19:04:39.000000000","message":"changed to, \"Size of WSGI thread pool\".","commit_id":"f2db135c9349706b4c85b9eb6b5bab51373773f9"},{"author":{"_account_id":748,"name":"Armando Migliaccio","email":"armamig@gmail.com","username":"armando-migliaccio"},"change_message_id":"ebe12218eb4d8ebe28e4014630875d91df1d0fab","unresolved":false,"context_lines":[{"line_number":55,"context_line":"    cfg.BoolOpt(\u0027use_ssl\u0027,"},{"line_number":56,"context_line":"                default\u003dFalse,"},{"line_number":57,"context_line":"                help\u003d_(\u0027Enable SSL on the API server\u0027)),"},{"line_number":58,"context_line":"    cfg.IntOpt(\u0027wsgi_default_pool_size\u0027,"},{"line_number":59,"context_line":"               default\u003d1000,"},{"line_number":60,"context_line":"               help\u003d_(\"Size of the pool of greenthreads used by wsgi\")),"},{"line_number":61,"context_line":""}],"source_content_type":"text/x-python","patch_set":6,"id":"dae33548_b1bc9e6d","line":58,"updated":"2016-02-15 17:25:42.000000000","message":"how are the other options like max_overflow etc tunable? They look hard-coded too in neutron/common/config.py","commit_id":"f7ef174e27b376574d9f2445c8f3c603e794c770"},{"author":{"_account_id":11816,"name":"mike bayer","display_name":"Mike Bayer","email":"mike_mp@zzzcomputing.com","username":"zzzeek","status":"Red Hat"},"change_message_id":"7677d2c23f9268d11f56835f52c239fc019c98b1","unresolved":false,"context_lines":[{"line_number":55,"context_line":"    cfg.BoolOpt(\u0027use_ssl\u0027,"},{"line_number":56,"context_line":"                default\u003dFalse,"},{"line_number":57,"context_line":"                help\u003d_(\u0027Enable SSL on the API server\u0027)),"},{"line_number":58,"context_line":"    cfg.IntOpt(\u0027wsgi_default_pool_size\u0027,"},{"line_number":59,"context_line":"               default\u003d1000,"},{"line_number":60,"context_line":"               help\u003d_(\"Size of the pool of greenthreads used by wsgi\")),"},{"line_number":61,"context_line":""}],"source_content_type":"text/x-python","patch_set":6,"id":"dae33548_9f8fe122","line":58,"in_reply_to":"dae33548_b1bc9e6d","updated":"2016-02-15 18:01:47.000000000","message":"those are just defaults.  the place where max_overflow is actually set states: \"These can be tweaked by the conf variables - max_pool_size, max_overflow and pool_timeout\"","commit_id":"f7ef174e27b376574d9f2445c8f3c603e794c770"},{"author":{"_account_id":9656,"name":"Ihar Hrachyshka","email":"ihrachys@redhat.com","username":"ihrachys","status":"Red Hat Networking Systems Engineer"},"change_message_id":"14346a11e911958d9166d1ad9fe2f1a7b1cd6357","unresolved":false,"context_lines":[{"line_number":55,"context_line":"    cfg.BoolOpt(\u0027use_ssl\u0027,"},{"line_number":56,"context_line":"                default\u003dFalse,"},{"line_number":57,"context_line":"                help\u003d_(\u0027Enable SSL on the API server\u0027))"},{"line_number":58,"context_line":""},{"line_number":59,"context_line":"]"},{"line_number":60,"context_line":""},{"line_number":61,"context_line":"CONF \u003d cfg.CONF"}],"source_content_type":"text/x-python","patch_set":9,"id":"da0c15f0_6751d062","line":58,"updated":"2016-03-24 23:15:12.000000000","message":"Unneeded change?","commit_id":"a984081d9d33476be92af432fa8f19b863607de5"},{"author":{"_account_id":11816,"name":"mike bayer","display_name":"Mike Bayer","email":"mike_mp@zzzcomputing.com","username":"zzzeek","status":"Red Hat"},"change_message_id":"6924f599049f747cde0739d817c6fb1b436321fb","unresolved":false,"context_lines":[{"line_number":55,"context_line":"    cfg.BoolOpt(\u0027use_ssl\u0027,"},{"line_number":56,"context_line":"                default\u003dFalse,"},{"line_number":57,"context_line":"                help\u003d_(\u0027Enable SSL on the API server\u0027))"},{"line_number":58,"context_line":""},{"line_number":59,"context_line":"]"},{"line_number":60,"context_line":""},{"line_number":61,"context_line":"CONF \u003d cfg.CONF"}],"source_content_type":"text/x-python","patch_set":9,"id":"9a061dce_b33c6ae3","line":58,"in_reply_to":"da0c15f0_6751d062","updated":"2016-04-06 02:34:58.000000000","message":"Done","commit_id":"a984081d9d33476be92af432fa8f19b863607de5"},{"author":{"_account_id":9656,"name":"Ihar Hrachyshka","email":"ihrachys@redhat.com","username":"ihrachys","status":"Red Hat Networking Systems Engineer"},"change_message_id":"14346a11e911958d9166d1ad9fe2f1a7b1cd6357","unresolved":false,"context_lines":[{"line_number":117,"context_line":"    def __init__(self, name, num_threads\u003d1000, disable_ssl\u003dFalse):"},{"line_number":118,"context_line":"        # Raise the default from 8192 to accommodate large tokens"},{"line_number":119,"context_line":"        eventlet.wsgi.MAX_HEADER_LINE \u003d CONF.max_header_line"},{"line_number":120,"context_line":"        self.num_threads \u003d num_threads or CONF.wsgi_default_pool_size"},{"line_number":121,"context_line":"        self.disable_ssl \u003d disable_ssl"},{"line_number":122,"context_line":"        # Pool for a greenthread in which wsgi server will be running"},{"line_number":123,"context_line":"        self.pool \u003d eventlet.GreenPool(1)"}],"source_content_type":"text/x-python","patch_set":9,"id":"da0c15f0_e7f2006f","line":120,"updated":"2016-03-24 23:15:12.000000000","message":"since in neutron/wsgi.py and in other places where we use the class we don\u0027t pass any value for num_threads, the config option won\u0027t ever be used, right?","commit_id":"a984081d9d33476be92af432fa8f19b863607de5"},{"author":{"_account_id":11816,"name":"mike bayer","display_name":"Mike Bayer","email":"mike_mp@zzzcomputing.com","username":"zzzeek","status":"Red Hat"},"change_message_id":"6924f599049f747cde0739d817c6fb1b436321fb","unresolved":false,"context_lines":[{"line_number":117,"context_line":"    def __init__(self, name, num_threads\u003d1000, disable_ssl\u003dFalse):"},{"line_number":118,"context_line":"        # Raise the default from 8192 to accommodate large tokens"},{"line_number":119,"context_line":"        eventlet.wsgi.MAX_HEADER_LINE \u003d CONF.max_header_line"},{"line_number":120,"context_line":"        self.num_threads \u003d num_threads or CONF.wsgi_default_pool_size"},{"line_number":121,"context_line":"        self.disable_ssl \u003d disable_ssl"},{"line_number":122,"context_line":"        # Pool for a greenthread in which wsgi server will be running"},{"line_number":123,"context_line":"        self.pool \u003d eventlet.GreenPool(1)"}],"source_content_type":"text/x-python","patch_set":9,"id":"9a061dce_933fa6d5","line":120,"in_reply_to":"da0c15f0_e7f2006f","updated":"2016-04-06 02:34:58.000000000","message":"that\u0027s supposed to be None, I think this got whacked in a merge.  However.   Do we want to assume CONF.wsgi_default_pool_size is 1000 outside, or should we assert in neutron that it defaults to 1000?  Beyond that, do we want to *change* the default to be something a little more useful, like 50?","commit_id":"a984081d9d33476be92af432fa8f19b863607de5"},{"author":{"_account_id":6524,"name":"Henry Gessau","email":"HenryG@gessau.net","username":"gessau"},"change_message_id":"ba4b6a2b3766b8d1130f16b06d5243219636acdc","unresolved":false,"context_lines":[{"line_number":116,"context_line":"    def __init__(self, name, num_threads\u003dNone, disable_ssl\u003dFalse):"},{"line_number":117,"context_line":"        # Raise the default from 8192 to accommodate large tokens"},{"line_number":118,"context_line":"        eventlet.wsgi.MAX_HEADER_LINE \u003d CONF.max_header_line"},{"line_number":119,"context_line":"        self.num_threads \u003d num_threads or CONF.wsgi_default_pool_size"},{"line_number":120,"context_line":"        self.disable_ssl \u003d disable_ssl"},{"line_number":121,"context_line":"        # Pool for a greenthread in which wsgi server will be running"},{"line_number":122,"context_line":"        self.pool \u003d eventlet.GreenPool(1)"}],"source_content_type":"text/x-python","patch_set":10,"id":"9a061dce_087d01b1","line":119,"updated":"2016-04-06 15:01:47.000000000","message":"If the user does not specify or configure a value, then it will default to 100 because of [1]. (Assuming I have followed the trail correcly.)\n\nAre we OK with this? What effect will it have on users when they upgrade and go from 1000 to 100?\n\n[1] https://github.com/openstack/oslo.service/blob/master/oslo_service/_options.py#L71-L73","commit_id":"d324248304109a04741885f9768f33c6019b5bd2"},{"author":{"_account_id":11816,"name":"mike bayer","display_name":"Mike Bayer","email":"mike_mp@zzzcomputing.com","username":"zzzeek","status":"Red Hat"},"change_message_id":"fc70a222512801649e126d1c715deb583a451c6c","unresolved":false,"context_lines":[{"line_number":116,"context_line":"    def __init__(self, name, num_threads\u003dNone, disable_ssl\u003dFalse):"},{"line_number":117,"context_line":"        # Raise the default from 8192 to accommodate large tokens"},{"line_number":118,"context_line":"        eventlet.wsgi.MAX_HEADER_LINE \u003d CONF.max_header_line"},{"line_number":119,"context_line":"        self.num_threads \u003d num_threads or CONF.wsgi_default_pool_size"},{"line_number":120,"context_line":"        self.disable_ssl \u003d disable_ssl"},{"line_number":121,"context_line":"        # Pool for a greenthread in which wsgi server will be running"},{"line_number":122,"context_line":"        self.pool \u003d eventlet.GreenPool(1)"}],"source_content_type":"text/x-python","patch_set":10,"id":"9a061dce_82178953","line":119,"in_reply_to":"9a061dce_087d01b1","updated":"2016-04-09 22:52:36.000000000","message":"So I\u0027d point you to the comment for that commit which is here: \n\nhttps://github.com/openstack/oslo.service/commit/321c185f3d977be153199a126720f6c3c00b0fdf\n\nit summarizes the testing that was done to come up with this number and also refers out to the full ML thread.\n\nBasic idea is, as long as most/all neutron web service requests use the database, and you\u0027re not setting your oslo.db connection pool to allow 1000 connections per process, then any number of greenlets over that max number (which defaults to 15) can\u0027t do any work anyway.   Better to have far fewer greenlets so that under high concurrency, requests are instead routed to other processes.","commit_id":"d324248304109a04741885f9768f33c6019b5bd2"},{"author":{"_account_id":6524,"name":"Henry Gessau","email":"HenryG@gessau.net","username":"gessau"},"change_message_id":"1e760d8645aa404478687a545eb24bea8a5006ec","unresolved":false,"context_lines":[{"line_number":116,"context_line":"    def __init__(self, name, num_threads\u003dNone, disable_ssl\u003dFalse):"},{"line_number":117,"context_line":"        # Raise the default from 8192 to accommodate large tokens"},{"line_number":118,"context_line":"        eventlet.wsgi.MAX_HEADER_LINE \u003d CONF.max_header_line"},{"line_number":119,"context_line":"        self.num_threads \u003d num_threads or CONF.wsgi_default_pool_size"},{"line_number":120,"context_line":"        self.disable_ssl \u003d disable_ssl"},{"line_number":121,"context_line":"        # Pool for a greenthread in which wsgi server will be running"},{"line_number":122,"context_line":"        self.pool \u003d eventlet.GreenPool(1)"}],"source_content_type":"text/x-python","patch_set":10,"id":"9a061dce_9d452819","line":119,"in_reply_to":"9a061dce_82178953","updated":"2016-04-10 00:09:46.000000000","message":"I am fine with this explanation for why it is a better number. It sounds like we can be reasonably confident that the change will not degrade anyone\u0027s installation when they upgrade.\n\nI still think the change in default needs to be mentioned in the release note.","commit_id":"d324248304109a04741885f9768f33c6019b5bd2"}],"releasenotes/notes/config-wsgi-pool-size-a4c06753b79fee6d.yaml":[{"author":{"_account_id":9515,"name":"Matthew Kassawara","email":"mkassawara@gmail.com","username":"ionosphere80"},"change_message_id":"4984ed24404214fc84f35ee01ff4d2a31392df46","unresolved":false,"context_lines":[{"line_number":1,"context_line":"---"},{"line_number":2,"context_line":"prelude: \u003e"},{"line_number":3,"context_line":"    The size of the pool of greenthreads used by wsgi"},{"line_number":4,"context_line":"    can now be configured."},{"line_number":5,"context_line":"other:"},{"line_number":6,"context_line":"  - Operators may want to tune the ``max_overflow`` and"},{"line_number":7,"context_line":"    ``wsgi_default_pool_size`` configuration options according"}],"source_content_type":"text/x-yaml","patch_set":6,"id":"dae33548_34dcfce3","line":4,"updated":"2016-02-15 17:43:57.000000000","message":"Suggestion:\n\nSupport configuration of greenthreads pool for WSGI.","commit_id":"f7ef174e27b376574d9f2445c8f3c603e794c770"},{"author":{"_account_id":11816,"name":"mike bayer","display_name":"Mike Bayer","email":"mike_mp@zzzcomputing.com","username":"zzzeek","status":"Red Hat"},"change_message_id":"ea55e3cb0fecb7d8d4a9f43a54573e340a0376a4","unresolved":false,"context_lines":[{"line_number":1,"context_line":"---"},{"line_number":2,"context_line":"prelude: \u003e"},{"line_number":3,"context_line":"    The size of the pool of greenthreads used by wsgi"},{"line_number":4,"context_line":"    can now be configured."},{"line_number":5,"context_line":"other:"},{"line_number":6,"context_line":"  - Operators may want to tune the ``max_overflow`` and"},{"line_number":7,"context_line":"    ``wsgi_default_pool_size`` configuration options according"}],"source_content_type":"text/x-yaml","patch_set":6,"id":"3afc51ec_6be7d8e4","line":4,"in_reply_to":"dae33548_34dcfce3","updated":"2016-03-11 19:04:07.000000000","message":"done","commit_id":"f7ef174e27b376574d9f2445c8f3c603e794c770"},{"author":{"_account_id":6524,"name":"Henry Gessau","email":"HenryG@gessau.net","username":"gessau"},"change_message_id":"88bef3a2e1bb6040700cd14db8fa165e84b793a4","unresolved":false,"context_lines":[{"line_number":3,"context_line":"    The size of the pool of greenthreads used by wsgi"},{"line_number":4,"context_line":"    can now be configured."},{"line_number":5,"context_line":"other:"},{"line_number":6,"context_line":"  - Operators may want to tune the ``max_overflow`` and"},{"line_number":7,"context_line":"    ``wsgi_default_pool_size`` configuration options according"},{"line_number":8,"context_line":"    to the investigations outlined in this `mailing list post"},{"line_number":9,"context_line":"    \u003chttp://lists.openstack.org/pipermail/openstack-dev/2015-December/082717.html\u003e`_."}],"source_content_type":"text/x-yaml","patch_set":6,"id":"dae33548_91a07af6","line":9,"range":{"start_line":6,"start_character":0,"end_line":9,"end_character":85},"updated":"2016-02-15 17:08:27.000000000","message":"I don\u0027t know if it is a good idea to link to the ML post here, but I just wanted to get the release note up. Now we can get opinions from reviewers.","commit_id":"f7ef174e27b376574d9f2445c8f3c603e794c770"},{"author":{"_account_id":748,"name":"Armando Migliaccio","email":"armamig@gmail.com","username":"armando-migliaccio"},"change_message_id":"ebe12218eb4d8ebe28e4014630875d91df1d0fab","unresolved":false,"context_lines":[{"line_number":3,"context_line":"    The size of the pool of greenthreads used by wsgi"},{"line_number":4,"context_line":"    can now be configured."},{"line_number":5,"context_line":"other:"},{"line_number":6,"context_line":"  - Operators may want to tune the ``max_overflow`` and"},{"line_number":7,"context_line":"    ``wsgi_default_pool_size`` configuration options according"},{"line_number":8,"context_line":"    to the investigations outlined in this `mailing list post"},{"line_number":9,"context_line":"    \u003chttp://lists.openstack.org/pipermail/openstack-dev/2015-December/082717.html\u003e`_."}],"source_content_type":"text/x-yaml","patch_set":6,"id":"dae33548_31408ea5","line":9,"range":{"start_line":6,"start_character":0,"end_line":9,"end_character":85},"in_reply_to":"dae33548_91a07af6","updated":"2016-02-15 17:25:42.000000000","message":"Unless you turn this into a blog post, I guess the link should stay. That said, this could capture the tl;dr version of the ML thread so that we can rid of the reference altogether.","commit_id":"f7ef174e27b376574d9f2445c8f3c603e794c770"},{"author":{"_account_id":6524,"name":"Henry Gessau","email":"HenryG@gessau.net","username":"gessau"},"change_message_id":"ba4b6a2b3766b8d1130f16b06d5243219636acdc","unresolved":false,"context_lines":[{"line_number":1,"context_line":"---"},{"line_number":2,"context_line":"prelude: \u003e"},{"line_number":3,"context_line":"    Support configuration of greenthreads pool for WSGI."},{"line_number":4,"context_line":"other:"}],"source_content_type":"text/x-yaml","patch_set":10,"id":"9a061dce_935dc27b","line":1,"updated":"2016-04-06 15:01:47.000000000","message":"Should we also mention that the default changes from 1000 to 100?","commit_id":"d324248304109a04741885f9768f33c6019b5bd2"},{"author":{"_account_id":6524,"name":"Henry Gessau","email":"HenryG@gessau.net","username":"gessau"},"change_message_id":"a6648b246174f4f246302c257292169044da1741","unresolved":false,"context_lines":[{"line_number":1,"context_line":"---"},{"line_number":2,"context_line":"prelude: \u003e"},{"line_number":3,"context_line":"    Support configuration of greenthreads pool for WSGI."},{"line_number":4,"context_line":"other:"}],"source_content_type":"text/x-yaml","patch_set":10,"id":"7a1b2928_ae371daf","line":1,"in_reply_to":"7a1b2928_4b9bdec9","updated":"2016-04-11 18:23:21.000000000","message":"Thanks!","commit_id":"d324248304109a04741885f9768f33c6019b5bd2"},{"author":{"_account_id":11816,"name":"mike bayer","display_name":"Mike Bayer","email":"mike_mp@zzzcomputing.com","username":"zzzeek","status":"Red Hat"},"change_message_id":"b8af8b84a043d99f4c732648fb95d3228d131b14","unresolved":false,"context_lines":[{"line_number":1,"context_line":"---"},{"line_number":2,"context_line":"prelude: \u003e"},{"line_number":3,"context_line":"    Support configuration of greenthreads pool for WSGI."},{"line_number":4,"context_line":"other:"}],"source_content_type":"text/x-yaml","patch_set":10,"id":"7a1b2928_4b9bdec9","line":1,"in_reply_to":"9a061dce_1d52b8dd","updated":"2016-04-11 16:59:29.000000000","message":"done, again ?","commit_id":"d324248304109a04741885f9768f33c6019b5bd2"},{"author":{"_account_id":6524,"name":"Henry Gessau","email":"HenryG@gessau.net","username":"gessau"},"change_message_id":"1e760d8645aa404478687a545eb24bea8a5006ec","unresolved":false,"context_lines":[{"line_number":1,"context_line":"---"},{"line_number":2,"context_line":"prelude: \u003e"},{"line_number":3,"context_line":"    Support configuration of greenthreads pool for WSGI."},{"line_number":4,"context_line":"other:"}],"source_content_type":"text/x-yaml","patch_set":10,"id":"9a061dce_1d52b8dd","line":1,"in_reply_to":"9a061dce_635a6052","updated":"2016-04-10 00:09:46.000000000","message":"Done where?","commit_id":"d324248304109a04741885f9768f33c6019b5bd2"},{"author":{"_account_id":20037,"name":"Petrônio Carlos Bezerra","email":"petroniocg@lsd.ufcg.edu.br","username":"petroniocg"},"change_message_id":"e5087b03eb17d22d34b86865dad3ec6bf5b9afbf","unresolved":false,"context_lines":[{"line_number":1,"context_line":"---"},{"line_number":2,"context_line":"prelude: \u003e"},{"line_number":3,"context_line":"    Support configuration of greenthreads pool for WSGI."},{"line_number":4,"context_line":"other:"}],"source_content_type":"text/x-yaml","patch_set":10,"id":"9a061dce_635a6052","line":1,"in_reply_to":"9a061dce_935dc27b","updated":"2016-04-08 13:37:34.000000000","message":"Done","commit_id":"d324248304109a04741885f9768f33c6019b5bd2"}]}
